In the current rapid development of CMOS image sensors, the image signal processing speed seriously hinders the development of high-pixel image sensors. Moreover, the image signal processing speed and the image signal gain also limits each other. In order to increase speed, the gain cannot be too high; but insufficient gain can in turn cause problems in low light image acquisition. Therefore, efforts to increase the gain while not impacting the processing speed has become a focal point in the competition among the design manufacturers of CMOS image sensors.
In the existing technology, the CMOS image sensor generally does not perform processing during image signal sampling operations. Thus the image signal directly samples to the capacitor and then the signal is directly read out. The signal margin only decreases and not increases. In order to solve the image sensor problem in sampling in a low light situation, the image signal sent out from the buffer undergoes a backend process. Thus the sampled signal is magnified; but the technical requirement for this process is much higher. Thus, the existing sampling control circuit structure remains simple and it does not have the ability to increase gain. This results in the output signal to be insufficient in gain, causing the low light effect to be less than ideal.